WO2019206867A1 - Method for detecting an operating anomaly of a battery and system implementing said method - Google Patents

Abstract

An aspect of the invention concerns a method (100) for detecting an anomaly in the operation of a battery using a battery management system, said system comprising an acoustic receiver configured to be attached to a wall of the battery, and a calculating means connected to the acoustic emitter and to the acoustic receiver, a mapping defining a first operating region termed the normal operating region, a second operating region termed the at-risk operating region, and a third operating region termed the dangerous operating region, said method comprising at least one first measurement cycle, each measurement cycle being separated from the preceding measurement cycle by a period termed the measurement period, each measurement cycle comprising a step (102) of receiving an acoustic signal by the acoustic receiver, the received signal being transmitted to the calculating means so as to obtain a measurement point in the mapping; a step (103) of determining the operating region in which the measurement point is located; and, when the measurement point is located in the at-risk operating region or in the dangerous operating region, a step (104) of detecting an anomaly.

Description

 METHOD FOR DETECTING AN OPERATING FAULT OF A BATTERY AND SYSTEM IMPLEMENTING SAID METHOD

TECHNICAL FIELD OF THE INVENTION

The technical field of the invention is that of the non-intrusive diagnosis and operation of electrochemical energy conversion systems (batteries, batteries, fuel cells). It relates more particularly to a method using ultrasonic characterization techniques to identify the optimal operating parameters of a battery in order to be able to detect a malfunction of the battery thus characterized and, possibly, make the appropriate corrections. The invention also relates to a system implementing said method.

BACKGROUND OF THE INVENTION

Battery management systems are essential in order to meet the demands of users as efficiently as possible. In general, the management systems take into account the state of charge of the battery, the state of health of the battery and / or the state of safety of the battery. This information is generally obtained by measuring electrical quantities at the battery terminals, but also by external data such as those provided by thermal analysis techniques (measurement of temperature and / or heat flux) or control. non-destructive (ultrasonic characterization for example). Once these parameters are determined, the management system regulates the operation of the battery to optimize performance.

The ultrasonic acoustic signal analysis techniques recently proposed for the study of operating batteries (Gold et al., 2017, Sood, Pecht, & Osterman, 2016, Steingart et al., 2016, DE102015210266A1), however, have certain drawbacks. Firstly, they are only interested in certain parameters of the battery and do not take into consideration all the information contained in the measured acoustic signal. In addition, the signals Acoustics used usually take the form of short-term pulses (less than 0.1 ms) that do not allow the establishment of a stationary regime within the battery, thus limiting the mathematical tools that it is possible to use to analyze said signals. Secondly, the techniques used do not take into account the peculiarities of each battery and do not allow to take into account the evolution of the characteristics of the battery during its life. Indeed, the main characteristics used for managing the operation of the batteries (minimum and maximum cell voltage values, maximum current, operating temperature range and storage ...) are generally provided by the manufacturer. Most of the time, they come only in the form of general guidelines on the limits for certain electrical characteristics, without taking into account the evolution of these limits during aging or other parameters of such as temperature.

There is therefore a need for a method for determining at any time the optimal parameters of use of a battery placed in different operating conditions and to take into account the evolution of the battery during its lifetime in order to detect a possible anomaly in the operation of the battery. SUMMARY OF THE INVENTION

The invention offers a solution to the problems mentioned above by allowing to take into account a plurality of parameters contained in the acoustic signals transmitted by the battery studied. From the analysis of these signals, it is thus possible to determine whether the battery is used in optimal conditions to guarantee its performance and durability, or conversely, if the conditions of use may be detrimental to its proper functioning . It is also possible to determine if the battery may pose a risk to the safety of the surrounding property and people. This allows a better diagnosis of the battery and thus to make a correction of the parameters of operation more efficiently and to maintain the battery in the best operating conditions.

One aspect of the invention relates to a method for detecting an abnormality in the operation of a battery using a battery management system, said system comprising an acoustic transmitter configured to be attached to a wall of the battery. battery, said acoustic transmitter being fixed on a wall of the battery during the implementation of the method, an acoustic receiver configured to be fixed on a wall of the battery, said acoustic receiver being fixed on a wall of the battery during the implementation of the method, and a calculation means connected to the acoustic receiver, a mapping defining a first operating zone called normal operating zone, a second operating zone called risk operating zone and a third operating zone called zone of operation. dangerous operation. The method according to a first aspect of the invention comprises at least a first measuring cycle, preferably a plurality of measurement cycles, each measuring cycle being separated from the preceding measurement cycle by a so-called measurement period, each cycle of measurement being measure comprising:

 a step of emitting an acoustic signal by the acoustic transmitter on instruction of the calculation means;

 - A step of receiving an acoustic signal by the acoustic receiver, the received signal being transmitted to the calculation means so as to obtain a measurement point in the map;

 a step of determining the operating zone in which the measurement point is located;

 - when the measuring point is located in the hazardous operation zone or in the hazardous operating zone, a step of detecting an anomaly.

Battery means any electrochemical system for converting and storing electrical energy (rechargeable or not). Thus, it is allowed, using at least one measurement cycle, to identify the operating mode of the battery and any anomalies. In addition, the step of transmitting an acoustic signal generating a response from the battery which is then received at the receiving step. It is therefore not necessary to continuously acquire the acoustic signal emitted by the battery, this acquisition being triggered simultaneously a transmission and over a period defined according to the duration of the transmitted signal. It also avoids noise filtering problems encountered when the acoustic signal is received is simply generated by the operation of the battery itself. Furthermore, the signal emitted during the transmission step has precisely defined properties (time interval between two transmissions, amplitude of the signal, its duration, frequency distribution of the signal), which facilitates the analysis of the signal received and allows a comparison of measurements over time. In addition, the use of operating zones makes it possible to limit the memory necessary for the implementation of the method compared to the techniques of the prior art. Indeed, in the techniques of the prior art, each measurement is compared to the measurements of a library which requires to keep in memory the large number of measures that constitute the library in question. In the method according to a first aspect of the invention, only operating areas are used, which requires much less memory. For example, a reference point can be memorized as well as the distance corresponding to the normality deviation which is admitted and which thus defines the limit between the normal operating zone and the risky operating zone and the boundary between the zone risky operation and the hazardous operating area. In addition, it is faster to locate a measuring point to identify in which of these three areas is said measuring point than to make a comparison with other points of a library as is done in the state of the art. It is important to note that a characteristic such as state of charge or state of health can not by itself be used to define a zone as described by the invention. Indeed, the same state of charge may correspond to operation in a "normal" zone, a "dangerous" zone or a "risk zone" depending on the temperature, the electrical power passing through the battery or the voltage of the battery. cell for example. Such situations arise in particular during "abusive" tests during which a battery is overcharged, overheated or short-circuited. In addition to the features that have just been mentioned in the preceding paragraph, the method according to a first aspect of the invention may have one or more additional characteristics among the following, considered individually or in any technically possible combination.

Advantageously, the duration of the tap is between 0.2 ms and 1 ms. Such a duration of the signal makes it possible to reduce the density of external noises likely to be superimposed on the signal and thus to mask the signals of short duration. It also makes it possible to establish a stationary regime in the battery, which is preferable when the acoustic signals are analyzed using mathematical tools such as the Fourier transform. This duration also makes it possible to reduce the amplitude of the signal emitted by the acoustic transmitter while guaranteeing the fact that the energy of the transmitted signal will be sufficiently high so that the latter is easily detected at the level of the acoustic receiver. Finally, this duration makes it possible to highlight the consequences of certain phenomena of partial transmission and reflection of the waves during the transport of the acoustic waves in the material and at the interfaces with the appearance of periodic modulation of the signal due to the superposition of signals. very close frequencies (also called beats).

Advantageously, the mapping comprises a reference point located in the normal operating zone, each zone is associated with a limit speed, a speed being positive when it moves away from the reference point and negative when it is close to the reference point. reference and each measurement cycle includes:

 a step of calculating the speed of the measurement point in the map, this speed being equal to the distance between the measurement point obtained during the preceding measurement cycle and the current measurement point (that is to say obtained in the current cycle) divided by the measurement period;

 a step of comparing the speed of the measurement point with the limit speed associated with the zone in which the measurement point is located;

when the speed of the measurement point is greater than the limit speed associated with the zone in which the second measurement point is located, the triggering of the step of detecting an anomaly. Thus it is possible to anticipate a degradation of the operating conditions of the battery through an evolution considered too fast the speed of the last measurement point. In addition, corrective action can be taken even before the battery enters a hazardous or hazardous operating area, contributing to longer battery life.

Advantageously, each measurement cycle comprises, when an anomaly is detected, a step of triggering an alert. Thus the user is warned in case of abnormal operation of the battery.

Advantageously, each cycle comprises, when an anomaly is detected or when an alert is triggered, a correction step of the battery operation regiment. Thus, in addition to alerting the user of a malfunction, the management system sets up corrective actions.

Advantageously, when the correction step does not make it possible to restore a normal operating mode of the battery after a predetermined time, a step of switching to degraded mode or a system shutdown step.

Thus, when the corrective actions implemented are not sufficient, the system ensures that the damage of the battery is limited by putting the battery in degraded mode or at a standstill.

Advantageously, a step of reducing the measurement period is triggered by the step of triggering an alert.

Thus, by reducing the measurement period, the method makes it possible to ensure better monitoring of the battery, the time separating two consecutive measurement cycles of the operating mode of said battery being reduced.

Advantageously, each measurement point obtained during each cycle is stored until a number of measurement points equal to a threshold value, said adaptation threshold, the position of the reference point is recalculated and the measuring points are erased. when said threshold is reached.

Thus, the mapping takes into account the new measurements which allows the system to perform a self-learning from said measurements. A second aspect of the invention relates to a battery management system comprising the means for implementing a method according to a first aspect of the invention.

A third aspect of the invention relates to a computer program comprising instructions that cause the management system according to a second aspect of the invention to perform the steps of the method according to a first aspect of the invention.

A fourth aspect of the invention relates to a computer readable medium on which the computer program according to a third aspect of the invention is stored.

The invention and its various applications will be better understood by reading the following description and examining the figures that accompany it.

BRIEF DESCRIPTION OF THE FIGURES

 The figures are presented as an indication and in no way limit the invention.

 - Figure 1 shows a schematic representation of a management system according to a second aspect of the invention.

 - Figure 2 shows a flow chart of a method according to a first aspect of the invention.

 FIG. 3 shows different signal patterns that can be used in a method according to a first aspect of the invention.

 - Figure 4 shows a map in which there are three areas corresponding to three different operating areas of the battery used in a method according to a first aspect of the invention.

FIGS. 5A and 5C show, for FIG. 5A, an example of an evolution of the absolute energy of the signal emitted in a method according to a first aspect of the invention as a function of the state of charge, for FIG. example of evolution of the absolute energy of the signal received in a method according to a first aspect of the invention as a function of the state of charge of the battery, and for FIG. 5C the evolution of the maximum recommended voltage at the terminals Lithium-ion battery type NMC / G (positive electrode based of metal oxide composed of nickel, manganese and cobalt, negative electrode based on graphite) as a function of temperature.

 FIG. 6 shows the evolution of the state of health of a battery with and without implementation of a method according to a first aspect of the invention.

 FIGS. 7A to 7C show, for FIG. 7A, the voltage at the terminals of the battery during a plurality of charge / discharge cycles carried out on a lithium-ion battery of the LFP / G type (phosphate-based positive electrode). graphite based negative electrode) for different temperatures, for FIG. 7B the evolution of the measured flight time on a LFP / G battery for different temperatures as a function of time, and for FIG. 7C the evolution of time of flight measured according to the temperature.

 FIGS. 8A to 8D show the analysis through different parameters of two different thermal behaviors of the same NMC / G type battery.

 FIGS. 9A to 9C illustrate the determination of the different operating zones of the same NMC / G type battery.

DETAILED DESCRIPTION OF AT LEAST ONE EMBODIMENT OF THE INVENTION

Unless otherwise specified, the same element appearing in different figures has a unique reference.

A first embodiment of a first aspect of the invention illustrated in FIGS. 1 to 4 relates to a method 100 for detecting an anomaly in the operation of a battery BAT using a battery management system. BAT. In the following, BAT battery means any electrochemical system for converting and storing electrical energy (rechargeable or not). Preferably, the battery management system BAT comprises means for controlling and measuring the reference physical quantities of the battery BAT such as the current delivered by the battery BAT or the voltage across the battery BAT. In addition, the management system comprises an acoustic receiver RA configured to be fixed on a wall of the battery BAT ( acoustic receiver being fixed on a wall of the battery during the implementation of the method according to a first aspect of the invention), and a calculation means MC connected to the acoustic receiver RA. The acoustic receiver RA is preferably a piezoelectric acoustic receiver RA and the latter is fixed to a wall of the battery BAT by means of an adhesive or any other fixing means. For example, the acoustic receiver RA may be based on PVDF (Vinylidene Polyfluoride Polymer), which device generally has broad resonance frequency bands between the kHz and the MHz. Alternatively, the acoustic receiver RA may be based on PZT ceramics (titanium lead zirconate), which device generally has narrow resonance frequency bands between 10 kHz and a few hundred kHz. The calculating means MC may take the form of a processor associated with a memory, an ASIC card or an FPGA. The connection of the calculation means MC with the acoustic receiver RA can be done by means of a bus, an Ethernet® type connection, or even by means of a wireless connection, for example Bluetooth®.

The management system also comprises, stored in a memory, a mapping CG defining a first operating zone called normal operating zone ZN, a second operating zone called risk operating zone ZR and a third operating zone called operating zone. dangerous ZD. More particularly, the normal operating zone ZN corresponds to a nominal operation of the battery BAT, the risk operating zone ZR corresponds to a risk operation of the battery BAT, that is to say the operating conditions in which the BAT battery can suffer irreversible damage in the short term if it is maintained beyond a predetermined duration, and the hazardous operating zone ZD corresponds to an operation resulting in an immediate irreversible degradation of the battery BAT. This mapping CG can be determined from a library of signals, each signal of the library of signals corresponding to a given operating mode of the battery BAT. For example, this library is made by learning. For this the BAT battery is placed in different operating conditions corresponding to the operating limits allowed (voltage, current, temperature). Then, the characteristics, preferably the main characteristics, of the acoustic signals received in the different normal operating conditions are stored in the library. Advantageously, it is also possible to store the evolution of the characteristics of the signals according to certain parameters (for example, the evolution of the signals as a function of the state of charge at several different temperatures: nominal, minimum, maximum temperature Operating). The library can for example be provided by BAT battery manufacturer. The characteristics of the signals may be derived from the temporal analysis of the signals (duration, absolute energy, signal strength, signal rise speed, flight time, maximum amplitude before duration, etc.) and / or frequency analysis (maximum frequency, average frequency, centroid frequency, power spectral density ...).

In an exemplary embodiment, the three operating zones are obtained from a library as described above. For this, the center of gravity of all the points of the library is calculated so as to obtain a reference point REF. Then, for each measuring point, the distance separating said point from the reference point REF is calculated so as to obtain the distribution of the distances. It is then possible to identify two distances defining two limits and thus three zones according to this distribution. For example, in a particularly advantageous mode, the library contains only measurements made in a normal operating regime. For each relevant parameter, a distance corresponding to approximately 95% of the measuring points (2s: 2 times the standard deviation of the deviations measured under normal operating conditions) is determined so as to define the radius of a sphere (the number of dimensions of this sphere is equal to that of the space in which is carried out the mapping CG) whose center is the reference point REF. This sphere then defines the boundary between the normal operating zone ZN and the risk operating zone ZR. Then a distance corresponding to 4 times the standard deviation (4s) and containing statistically more than 99.99% of the measurement points obtained under normal conditions, is determined so as to define the radius of a second sphere whose center is the point reference REF. This second sphere then defines the boundary between the ZR risk area and the ZD hazardous area. This embodiment is particularly advantageous insofar as it makes it possible to determine the three zones while remaining in a normal operating regime. Indeed, in this embodiment, the term "normal" operating zone means an area in which the battery is in operation (charge-discharge) or at rest in operating conditions defined as "normal" by the manufacturer. That is, it does not come out of nominal temperature, cell voltage and current (or power) ranges, usually detailed in the data sheet provided by the manufacturer. In addition, the physical characteristics of the acoustic waves received after each emission are within the limits defined during calibration and operation previously considered "normal" (ie in the limit of 2s presented above). Similarly, in this embodiment, the term "at risk" operating zone means an area in which the battery is still in operation (charge-discharge) or at rest in operating conditions defined as "normal" by the manufacturer. That is, it does not come out of nominal temperature, cell voltage and current (or power) ranges, usually detailed in the data sheet provided by the manufacturer. But this time, the physical peculiarities of the acoustic waves received after each emission go beyond the limits defined during the calibration and operation previously considered as "normal" (that is to say within the limit of 4s presented above). In addition, in this embodiment, the term "dangerous" operating zone means a zone in which the battery is still in operation (charge-discharge) or at rest under operating conditions defined as "normal" by the manufacturer. That is, it does not come out of nominal temperature, cell voltage and current (or power) ranges, usually detailed in the data sheet provided by the manufacturer. But this time, the physical peculiarities of the acoustic waves received after each emission go beyond the limits defined during the calibration and operation previously considered "at risk" (that is to say beyond the 4s limit presented above). Preferably, only the main characteristics of the signals, ie the most relevant characteristics to be taken into account, are used for the generation of the CG cartography. However, these characteristics depend strongly on the complete system and the mapping is therefore specific to said system. More particularly, the information needed to feed the library depends heavily on the complete system set up and in particular the following parameters:

 - the acoustic transmitter EA (positioning relative to the battery, signal form frequency range, amplitude ...);

 - particularities of BAT battery (size, shape, chemistry, packaging);

 particularities of the piezoelectric elements used for recording the received signals (resonance frequency, sensitivity, positioning and quality of the coupling with the battery, etc.).

The method 100 according to a first aspect of the invention comprises at least a first measurement cycle, preferably a plurality of measurement cycles, each measurement cycle being separated from the previous (or next) measurement cycle by a so-called period of measurement. measured. In other words, BAT battery monitoring is ensured by the succession in the time of measuring cycles. Preferably, this measurement period is constant. However, the latter may have to evolve, for example this period may be reduced if the evolution of the battery BAT operating regime leaves fear of damage to the latter, so as to ensure more frequent monitoring.

For this, each measurement cycle comprises a step 102 for receiving an acoustic signal by the acoustic receiver RA, the received signal being transmitted to the calculation means MC so as to obtain a measurement point in the CG map. This acoustic signal can be produced by an external source of acoustic signals (noise) capable of generating acoustic waves in the desired frequency range. This external source may for example be a processor, a fan or a switching power supply based on transistors. In the case of using an external source of acoustic signal, it is necessary to have at least one acoustic receiver RA but it is preferable to have at least two acoustic receivers RA. Indeed, with several acoustic receivers RA, the measurements to be considered for the analysis may also include parameters concerning the differences between the transmitted signal and the signals received by each of the RA receivers.

Preferably, the transmission of the received signal is controlled. For this, in one embodiment, the management system comprises an acoustic transmitter EA configured to be fixed on a wall of the battery BAT, said acoustic transmitter being fixed on a wall of the battery during the implementation of the method according to a first aspect of the invention, the calculating means MC being connected to the acoustic transmitter EA. The acoustic transmitter EA is preferably a piezoelectric acoustic transmitter EA and the latter is fixed to a wall of the battery BAT by means of an adhesive or any other means of attachment. For example, the acoustic emitter EA may be based on PVDF (Vinylidene Polyfluoride Polymer), a device which generally has broad resonance frequency bands between the kHz and the MHz. Alternatively, the acoustic transmitter EA may be based on ceramics PZT (lead titano-zirconate), a device which generally has narrow resonance frequency bands between 10 kHz and a few hundred kHz. Preferably, the wall on which is fixed the acoustic transmitter EA and opposite the wall on which is fixed the acoustic receiver RA. In addition, each measurement cycle comprises, before the step 102 of receiving the acoustic signal, a step 101 of emission of an acoustic signal by the acoustic transmitter EA on instruction of the calculation means MC. As already mentioned, the emission of a controlled acoustic signal has several advantages. First, it is not necessary to continuously acquire the acoustic signal emitted by the battery, this acquisition being done only when there is transmission. It also avoids noise filtering problems encountered when the acoustic signal is received is simply generated by the operation of the battery itself. Moreover, the signal emitted during the transmission step has well-known properties, which facilitates the analysis of the received signal and allows a comparison of the measurements over time.

Preferably, the transmitted signal comprises a plurality of frequencies between 1 kHz and 1 MHz, preferably between 100 kHz and 200 kHz. It is thus possible to excite the acoustic transmitter EA over part or all of this frequency range. The signal may in particular comprise several successive frequency components (for example, several sinusoids of increasing or decreasing frequency successively in time) and / or simultaneous (multi frequency excitation). As illustrated in FIG. 3, the pattern of the transmitted signal may be a square pattern, a triangle pattern, a sinusoidal pattern, a combination of at least part of these patterns, or even any shape. The shape, frequency and intensity of the signals may be chosen so as to take into account the specificities of each type of BAT battery, the nature of the acoustic transmitter EA and / or the nature of the acoustic receiver RA.

Preferably, the duration of the transmitted acoustic signal is between 0.2 and 1 ms. This duration makes it possible to reduce the density of external noises likely to be superimposed on the signal and thus to mask the signals of short duration. It also makes it possible to establish a stationary regime in the BAT battery, which is preferable when the acoustic signals are analyzed using mathematical tools such as the Fourier transform. It is useful to emphasize that such a stationary regime can not be obtained with short-term pulses (typically less than 0.1 ms) as used in prior art methods. This duration also makes it possible to reduce the amplitude of the signal emitted by the acoustic emitter EA while ensuring that the energy of the signal is sufficiently high so that the latter is easily detected at the level of the acoustic receiver RA. The reduction of the amplitude (compared to a short acoustic signal) makes it possible to reduce the problems related to the non-linearities of the response of the battery BAT, the acoustic transmitter EA and / or the acoustic receiver RA. In this respect, it is interesting to note that the acoustic signals detected by the acoustic receiver RA are converted into electrical signals so that they can be sent and processed by the calculation means MC. This conversion is reliable only if the non-linear phenomena are reduced to a minimum, which is made possible in the present invention by the duration of the acoustic signals used. In addition, the signals received by the acoustic receiver RA are obtained from a steady-state system, a condition necessary for the exploitation of said signals by data processing techniques. such as the Fourier transform, where the methods of the state of the art are confined to the techniques of temporal data processing. Finally, this duration makes it possible to highlight the consequences of certain phenomena (partial transmission and reflection of the waves during the transport of acoustic waves in matter and interfaces) related to the transport of acoustic waves in matter through the appearance of periodic modulation of the signal due to the superimposition of very close frequency signals (also called beats). But this type of signal is very rich in information related to the transport of acoustic waves through the different materials and the interfaces that make up the battery BAT.

Each measurement cycle also comprises a step 103 for determining the operating zone in which the measurement point is located. As previously explained, the calculating means MC has in memory a mapping CG comprising three zones. Once a measurement point is obtained, it is possible to determine in which of these three areas of this map CG it is located. In addition, each measurement cycle comprises, when the measuring point is in the risk operation zone ZR or in the hazardous operation zone ZD, a step 104 for detecting an anomaly.

To illustrate the influence of the operating conditions of a battery BAT on the signal measured during the implementation of the method 100 according to a first aspect of the invention, various experimental measurements have been made and are presented in FIGS. .

FIG. 5A shows the evolution of the signal strength as a function of time for three successive charge-discharge cycles C1, C2, C3. The first cycle C1 charge / discharge is a cycle for which the charge is complete while the second cycle C2 charge / discharge is a cycle for which the battery is charged to 1 10% of its nominal capacity. Cycle C3 is again a charge-discharge cycle identical to C1. For each cycle C1, C2, C3 of charge / discharge, two peaks of maximum force are present, between which is a plateau PL for which the signal strength (or energy of the signal) is substantially constant. However, we observe that the ordinate of this plateau in overload case is different from the ordinate of said plateau in case of normal load, this difference being materialized by DR in Figure 5A. FIG. 5B illustrates the evolution of the absolute energy of the signal received as a function of the state of charge of the battery during these three successive cycles C1, C2, C3 of charge-discharge. In the measurement library, the measurements associated with the first plate PL would therefore be associated with a normal charge of the battery BAT while the measurements associated with the second plate PL would be associated with an overload. The operating zones will then be plotted accordingly so that the measuring points corresponding to the first cycle C1 and the third cycle C3 are in the normal operating zone ZN while the measurement points corresponding to the overload during the second cycle are located in the risk zone ZR or in the danger zone ZD.

FIG. 5C shows the evolution of the maximum voltage across the battery BAT before reaching the risk zone ZR as a function of the temperature. Passage into the risk zone ZR is observed slightly below 4.2 V at 5 ° C and the voltage value decreases as the temperature increases. Thus, at 25 ° C, the maximum voltage before entering the ZR risk zone is only 4.05 V. It will be only 3.97 V for an operating temperature of 45 ° C. It is thus possible to use the detection of the passage in the risk zone ZR to determine the maximum operating voltage of the battery BAT and improve the management of this battery BAT to guarantee a better durability. This is illustrated in particular in FIG. 6, which presents an example of the evolution of the state of health of two BAT batteries tested with an accelerated aging protocol. The state of health of the two batteries BAT decreases during cycling but the one that benefits from optimized management by detecting the passage in the ZR risk zone has a deterioration of the state of health almost twice as low.

Another example of the influence that temperature can have on the operating speed of the battery is given in FIG. 7A, which shows a plurality of charge / discharge cycles carried out for different temperatures. More in particular, it presents on the one hand the evolution of the voltage across the battery BAT as a function of time (the black curve) as well as the evolution of the temperature of said battery BAT as a function of time (the gray curve) . Among the various parameters characterizing the acoustic signals transmitted, some do not evolve, or very little with the state of charge of the battery, but on the other hand, change significantly with the temperature. In the example described (measurements of acoustic signals made during charge-discharge cycles of a cylindrical battery LFP / G during temperature variations), this is the case of the duration of the received signal, the number of strokes, or, to a lesser extent, the average frequency of the signals. This is also the case of the flight time, that is to say the time separating the transmission of the transmitted signal from its reception.

FIGS. 7B and 7C show, for FIG. 7B, the flight time (the measuring points) and the temperature (the black curve) as a function of time and for FIG. 7C the flight time as a function of the temperature (the points represent the measures and the dashed line the linear regression). From these two figures, it is therefore possible to confirm that the operating temperature is strongly correlated with the measured flight time. It is therefore possible to take into account, in the method according to a first aspect of the invention, the operating temperature of the battery BAT by the measurement of the flight time. This consideration of variation of the internal temperature of the battery BAT can be used in addition to measurements by temperature sensors to detect unusual local heating and to prevent risks of fire or explosion. The internal temperature measurements allow a much earlier detection and with a higher accuracy of heating with the use of temperature sensors placed on the surface of the cells of the batteries and measuring only a "skin temperature" .

Yet another example of the influence of temperature is illustrated in Figures 8A-8D. FIG. 8A shows the evolution of the voltage (highest curve) and of the heat flux released during the charging of the same battery BAT in two distinct cases: firstly, in the case of an operation where the thermal behavior of the BAT battery is normal (lowest curve) with a moderate heat release, culminating at around 150 mW at the end of charging; then in a case where (middle curve), in exactly the same operating conditions, the heat flow generated by the battery BAT increases very abnormally to 650 mW. Such a quantity of energy released in the form of heat is the sign of a major dysfunction within the battery BAT, linked to the appearance of parasitic reactions. The immediate consequences may be damage to the materials, leading to irreversible degradation of performance and a significant risk of thermal runaway that can lead to BAT battery destruction as well as safety issues (gassing, the safety vent, fire, explosion ...). It is therefore essential to be able to detect such a malfunction as soon as possible to carry out the appropriate corrective actions (current limitation, emergency stop, etc.).

FIG. 8B shows the evolution of the voltage at the terminals of the battery BAT during charging in both cases (normal and abnormal operation) and illustrates that the evolution of the voltage across the battery BAT does not does not detect this abnormal heating. FIG. 8B also shows the evolution of the surface temperature of the battery BAT in both cases (normal and abnormal behavior). The abnormal heat release is well detected, but in this case only increases by a few degrees the surface temperature (5 ° C) compared to normal operation. Such heating can easily go unnoticed while its consequences can be very detrimental to subsequent performance or even the security of the system.

FIG. 8C shows the evolution of the absolute energy of the signal as a function of time for the normal and abnormal behavior and reveals that the abnormal thermal behavior is detectable by the measurement of the absolute energy of the acoustic signals: in fact the latter , in the case of abnormal behavior, no longer follows the normal evolution, the absolute energy of the acoustic signals decreasing sharply. As illustrated in FIG. 8D, a same measurement of the absolute energy can be made as a function of the state of charge and also reveals that abnormal behavior can be demonstrated by such a measurement. From these different measurements, it is possible to determine a map of BAT battery operation. FIG. 9A presents a simplified cartography (the latter taking into account only two parameters) presenting on the abscissa the difference at the centroid frequency and on the ordinate the difference with the absolute energy. Each measurement point in this space corresponds to a received acoustic signal. The dots in white correspond to operation in normal conditions, the dots in black correspond to the acoustic signals recorded when the battery BAT has an abnormal heat release during its operation.

Three operating zones have been defined in this space: ZN normal operating zone, ZR hazardous operating zone and ZD hazardous operating zone. In one embodiment, these three zones are defined from the distribution of the measurements obtained for the parameters considered (FIGS. 9B and 9C). A statistical analysis calculates the standard deviation of the values measured under normal operating conditions for each parameter. Thus, for each parameter, the limit between the normal operating zone ZN and the zone presenting a risk ZR can be defined from the measured standard deviations. In one embodiment, the tolerated deviation for the zone ZN is 2s (contains statistically more than 95% of the values obtained in normal operation) and the limit defined for the risk zone can be calculated from a difference of 4s (containing statistically more than 99.99% of the values obtained in normal operation). The tolerated difference may of course be adapted according to the intended use. It appears clearly on this map that the measurements corresponding to an abnormal regime (black dot) leave the normal operating zone ZN to reach the risk operating zone ZR and even the dangerous operating zone ZD. The abnormal behavior can therefore be detected using a method 100 according to a first aspect of the invention.

These various examples make it possible to show how the taking into account of a plurality of parameters of the measured signal makes it possible to detect one or more anomalies (overload, heating, etc.), where the measurement of a parameter single (state of charge, voltage across the battery BAT, etc.) can not be sufficient.

In one embodiment, the CG map comprises a reference point REF located in the normal operating zone ZN. In addition, each zone is associated with a speed limit. The velocity is positive when it moves away from the reference point REF and negative when it approaches the REF reference point. In addition, each measuring cycle comprises a step of calculating the speed of the measuring point in the CG map, this speed being equal to the distance between the measurement point obtained during the previous measurement cycle and the current measurement point ( that is, corresponding to the current measurement cycle) divided by the measurement period. It is useful to note that each received signal has a large number of parameters (frequencies, amplitudes, forces, energies, durations, flight times, etc.), and therefore CG mapping is performed in a large space in which each measurement obtained during a measurement cycle is a point in said space. It is thus possible to compare the different points of this space, in particular by measuring the distance separating two points. This distance can for example be Euclidean, Manhattan or Minkowski distance.

It also comprises a step of comparing the speed of the measurement point with the limit speed associated with the zone in which the measurement point is located. Finally, when the speed of the measurement point is greater than the limit speed associated with the zone in which the measurement point is located, it finally comprises triggering the step 104 for detecting an anomaly. Note that the comparison takes into account the sign of speed. Thus, a measuring point which tends to approach the reference point, and therefore of the normal operating zone, will not give rise to the detection of an anomaly if the measuring point itself is not located. outside the normal operating zone ZN.

In one embodiment, each measurement cycle comprises a step 105 of triggering an alert when an anomaly is detected. Thus, the method makes it possible to identify when the operating mode of the battery is abnormal. In one embodiment, a step of correcting the battery operation regiment BAT is triggered by step 105 of triggering an alert. Thus, in addition to notifying the user of the presence of a problem in BAT battery operation, the management system may initiate one or more corrective actions. Corrective actions include limiting the maximum power delivered by the BAT battery, reducing the maximum voltage range allowed across the BAT battery, or even limiting the acceptable temperature range for the BAT battery. In one embodiment, the corrective action or actions can be performed in a gradual manner, preferably by performing a measurement cycle between each correction to measure the effect of said correction, so as to limit the impact of the correction. or corrective actions on BAT battery operation. In one embodiment, the method 100 comprises, when the correction step does not make it possible to restore a normal operating mode of the battery BAT after a predetermined time (or after a predetermined number of measurement cycles), a step of switching to a degraded mode or a step of stopping the system.

In one embodiment, a step of reducing the measurement period is triggered by the step 104 of triggering an alert. Thus, when the battery BAT is in an abnormal operating mode, the regime of the latter is measured more regularly so that any corrective actions, the transition to degraded mode and / or the shutdown of the system are triggered as quickly as necessary. In one embodiment, when the operating mode of the battery BAT returns to normal, the measurement period can be reset to its initial value.

In one embodiment, each measurement point obtained during each cycle is stored until a number of measurement points equal to a threshold value is obtained (for example, a threshold value greater than or equal to 100 so as to obtain a representative statistical distribution), said adaptation threshold, the position of the REF reference point being recalculated then the measurement points erased when said threshold is reached. The storage of these measurement points can be done in the memory of the calculation means MC, the part of the memory thus occupied being released when the position of the reference point REF is recalculated. Preferably, the position of the new reference point REF corresponds to the centroid of the stored measurement points. This makes it possible to modify the position of the reference point REF during the life of the battery BAT and thus to take into account its aging. It should be noted that the change of position of the reference point also changes the three operating zones described previously.

In order to implement a method 100 according to a first aspect of the invention, a second aspect of the invention relates to a BAT battery management system. The battery management system BAT may comprise means for controlling and measuring the reference physical quantities of the BAT battery such as the current delivered by said BAT battery or the voltage across said BAT battery. These control means may in particular be used in order to correct the operating mode of the battery BAT if such a correction is necessary. In addition, the battery management system BAT comprises an acoustic transmitter EA configured to be fixed on a wall of the battery BAT. The acoustic transmitter EA is preferably a piezoelectric acoustic transmitter EA and the latter is fixed to a wall of the battery BAT by means of an adhesive or any other means of attachment. The BAT battery management system also includes an acoustic receiver RA configured to be attached to a wall of the BAT battery. As with the acoustic transmitter EA, the acoustic receiver RA is preferably a piezoelectric acoustic receiver RA and the latter is fixed to a wall of the battery BAT by means of an adhesive or other means of attachment. For example, the EA / RA acoustic transceiver may be based on PVDF (Vinylidene Polyfluoride Polymer), which device generally has broad resonance frequency bands between the kHz and the MHz. Alternatively, the acoustic transmitter / receiver EA / RA may be based on PZT (lead titano-zirconate), which device generally has narrow resonance frequency bands between 10 kHz and a few hundred kHz. Preferably, the wall on which is fixed the acoustic transmitter EA and opposite the wall on which is fixed the acoustic receiver RA. The management system also comprises a calculation means MC connected to the acoustic transmitter EA and the acoustic receiver RA. The computing means MC can take the form of a processor associated with a memory, an ASIC card or an FPGA. The connection of the calculation means MC with the acoustic transmitter and the acoustic receiver can be done by means of a bus, an Ethernet® type connection, or even by means of a connection in wires, for example Bluetooth ®. The calculating means MC also includes a memory in which a map CG is stored. This mapping CG defines a first operating zone called normal operating zone ZN, a second operating zone called risk operating zone ZR and a third operating zone called hazardous operating zone ZD.

In addition, the acoustic transmitter EA is configured to transmit on instruction of the calculation means MC an acoustic signal by the acoustic transmitter. Preferably, the acoustic signal comprises a plurality of frequencies between 1 kHz and 1 MHz, preferably between 100 kHz and 200 kHz. Preferably, the acoustic signal has a duration of between 0.2 and 1 ms. In addition, the acoustic receiver RA is configured to receive the acoustic signal and to transmit said signal to the calculation means MC. Finally, the calculation means MC is configured to determine the operating zone in which the measuring point is located and, when the measuring point is in the risk operation zone ZR or the hazardous operating zone ZD, detecting an anomaly.

Claims

A method (100) for detecting an abnormality in the operation of a battery (BAT) using a battery management system (BAT), said system comprising an acoustic transmitter (EA) configured to be fixed on a wall of the battery (BAT), an acoustic receiver (RA) configured to be fixed on a wall of the battery (BAT), said acoustic receiver being fixed on a wall of the battery during the implementation of the method , and a calculation means (MC) connected to the acoustic transmitter (EA) and to the acoustic receiver (RA), a mapping (CG) defining a first operating zone called normal operating zone (ZN), a second operating zone said operation zone at risk (ZR) and a third operating zone called hazardous operating zone (ZD), said method (100) comprising at least a first measurement cycle, each measurement cycle being separated from the previous measurement cycleby a so-called measurement period, each measurement cycle comprising:  a step (101) of emission of an acoustic signal by the acoustic transmitter (EA) on instruction of the calculation means (MC); - A step (102) for receiving an acoustic signal by the acoustic receiver (RA), the received signal being transmitted to the calculation means (MC) so as to obtain a measurement point in the map (CG);  a step (103) of determining the operating zone in which the measurement point is located;  - when the measuring point is located in the hazardous operation zone (ZR) or in the hazardous operating zone (ZD), a step (104) for detecting an anomaly. 2. Method (100) according to the preceding claim wherein the map comprises a reference point located in the normal operating zone (ZN), each zone is associated with a limit speed, a speed being positive when moving away from the reference point (REF) and negative when approaching the reference point (REF), each measurement cycle comprising:  a step of calculating the speed of the measurement point in the map, this speed being equal to the distance between the measurement point obtained during the previous measurement cycle and the current measurement point divided by the measurement period;  a step of comparing the speed of the measurement point with the limit speed associated with the operating zone in which the measurement point is located;  when the speed of the measurement point is greater than the limit speed associated with the operating zone in which the second measuring point is located, the triggering of the step (104) for detecting an anomaly. 3. Method (100) according to one of the preceding claims wherein, each measurement cycle comprises, when an abnormality is detected, a step (105) for triggering an alert. 4. Method (100) according to one of the preceding claims wherein each cycle comprises, when an abnormality is detected or an alert is triggered, a step of correcting the operating mode of the battery (BAT). 5. Method (100) according to the preceding claim comprising, when the correction step does not restore a normal operating mode of the battery (BAT) after a predetermined time, a step of switching to degraded mode or a step of stopping the system. 6. Method (100) according to one of the three preceding claims wherein a step of reducing the measurement period is triggered by the step (104) of triggering an alert. 7. Method (100) according to one of the preceding claims and claim 3 wherein each measuring point obtained during each cycle is stored until a number of measurement points equal to a threshold value, said adaptation threshold, the position of the reference point (REF) is recalculated and the measurement points erased when said threshold is reached. 8. Battery management system comprising the means for implementing a method (100) according to one of the preceding claims. 9. Computer program comprising instructions that lead the management system according to the preceding claim to perform the steps of the method (100) according to one of claims 1 to 7. 10. Computer-readable medium on which the computer program according to claim 9 is recorded.